Data Methodologies & Sources

Notice: I am aware of a couple major issues with the data I used to seed this site. See Known Issues below for details.

The historical college football data displayed on this website is compiled, cleaned, and maintained using a multi-layered approach. Because college football history spans over a century, records can often be conflicting, fragmented, or incomplete.

To ensure the highest possible accuracy, I combined automated data pipelines with meticulous archival research and manual verification. Below is an overview of the core sources and methodologies used to build our database.

Core Data Sources

My database is built upon the foundational work of several incredible sports data projects and archives:

  • CollegeFootballData.com (CFBD API): The primary baseline for our modern and historical data structures was built using the open-source API provided by CollegeFootballData.com. This robust framework provided our initial sets of teams, schedules, and historical game outcomes.
  • College Football Data Warehouse (via the Internet Archive): To enrich my historical records (particularly for early-era teams and defunct programs), I heavily leveraged data from College Football Data Warehouse. As the original site is no longer active, I meticulously scraped team-specific pages using the Internet Archive's Wayback Machine.
  • Wikipedia & Public Domain Archives: For gaps in early-century records, missing game details, or school-specific timeline anomalies, I utilized public domain information, primarily sourced from Wikipedia's extensive college football archives and verified university athletics histories.

Data Methodology & Enrichment Pipeline

Raw data only tells part of the story. To turn these disparate sources into a cohesive relational database, I implemented a rigorous enrichment process:

  1. Database Architecture & Integration: Initial datasets were normalized and ingested into a custom database. I mapped historical team names, conference realignments, and shifting school identities across a unified timeline to ensure cross-era queries remain accurate.
  2. Anomaly Detection & Quality Control: Once the foundational data was merged, I ran targeted diagnostic scripts to identify historical anomalies. This included searching for:
    • Duplicate records for the same game.
    • Discrepancies in total games played per season.
    • Games present from College Football Data Warehouse but not present from the CollegeFootballData.com or vice versa.
    • Inconsistent dates or missing venue information for early-era matchups.
  3. Manual Verification & Fact-Checking: Every flagged anomaly underwent manual research. I used targeted cross-referencing via search engines to find primary source materials, contemporary newspaper clippings, and official university records to resolve conflicting data points and fill in missing historical details.

Known Issues

The following are the known issues currently for this site:

  • There are large gaps of missing games with much of the earliest years for many teams. I haven't been able to pinpoint a year where data seems to start missing, but from the teams I have investigated, the years tend to be pre-1950.
  • The later games (post about 2005) are listed with exact start times of the games. The earlier games (pre-2005) are listed with only the date. Since I assumed all games had exact start times and convert to just the date, this resulted in games being listed as occurring on a different date than the actual game.

I am actively working to address these issues. The process has proven to be a bit tedious since I am using a separate data set to compare what I have seeded to the new data set and the new data set refers to teams by different names. Therefore, I need to map the names from the new data set to the names used in the existing data set. I will work through this data starting with the FBS and working my way through the other divisions of college football and post the updated data when I am done each division. I hope to get the FBS fixed sometime earlier in the 2026 season. The rest of the divisions may need to wait until after the 2026 season.

Disclaimer & Community Corrections

Historical data tracking is an evolving project. While I strive for absolute accuracy, errors can persist due to conflicting official records from the early decades of the sport. If you spot an error, an incorrect score, or a missing data point, I welcome your feedback. Please see my Contact Page for instructions on how to suggest an update.